LM5157-Q1: Sepic converter design

Hi Bryan,

But my customer wants to know how to calculate a value, not just a calculation tool.

Hi Lin,

Thanks for the feedback. I've found an application note that goes into the basic calculations necessary for SEPIC design. Furthermore, this is the user's guide for the LM5157 SEPIC evaluation module.

If you have certain specific values in mind that you'd like to receive calculation help with, please let me know.

Thank you and best regards,
Bryan

Hi Bryan,

Thanks a lot.

Can you tell me how fbw is produced? Is it frhpz/5 ?

And at the COMP,I can't know how to design value of R and C.

Hi Lin,

You're very welcome. Could you please attach the filled out quickstart calculator, with these values I can help you design and explain the various parts of the circuit.

Thank you and best regards,
Bryan

Hi Bryan,

Here is the data.

Please help me to know how to set the value of Fco,Rcomp,Ccomp and Chf.

Hi Bryan,

Perhaps using the frhpz value setting in the converter for bandwidth might be a better approach, but what is the reason for setting the RC compensation in the loop? What impact will not setting it have on the circuit?

Because I found the calculation formula for RC loop compensation in this document: https://www.ti.com/lit/an/snva168e/snva168e.pdf?ts=1700269847704 However, I'm unsure about the issues that might arise from not configuring it.

Hi Lin,

setting the RC on the COMP pin configures the poles and zeros for the feedback loop. This, when properly done, ensure the stability of the power stage.

Otherwise the power stage might be not stable, can start to oscillate and worst case get damaged.

See also:

Switch-mode power converter compensatin made easy

Best regards,

 Stefan

Hi Stefan,

OK,So To ensure circuit stability, I need to configure Rcomp, Ccomp, and Chf.

Would different converters have different solving formulas?

Do I simply need to follow the formulas provided in the 'How to Design a Boost Converter Using LM5157x, LM5158x' document to calculate Rcomp, Ccomp, and Chf?

Hi Lin,

Thanks for the reply. It would be a good assumption to start using the values that are calculated in the Quickstart Calculator.
Otherwise you could use the documentation that TI has published for the LM5157-Q1.

But like I said before, these are purely theoretical values and the actual tuning should be done in the lab, since the component placement and PCB routing influences the actual behavior of the system.

Thank you and best regards,
Bryan

Hi LIn,

there is a good application report about compensation. I think this will help to answer your questions:

Switch-mode power converter compensatin made easy

Best regards,

 Stefan

Hi Stefan,

There are recommended bandwidth values ​​in Excel. Do I need to set the values ​​smaller than each other?

In addition, should the value of compensation RC be set smaller than the recommended value?

Finally, could you please tell me how to verify that the bandwidth meets the design value through circuit simulation, and is the compensation RC setting correct?

Hi Lin,

if you study the document I suggested above, you will see that for stability a Phase and Gain Margin should be ensured.

There are some variance on what is recommended but this is also something the user can decide how the environment and what margin he would like to set.

We recommend a phase Margin of > 60 Degree and a gain margin of at least 12dB.

The suggested values are just recommendations. You can adjust them and see the result in the graph on the right side.

To check this, the best is to measure on the real board with a vector network analyzer as this will also capture all the parasitic of the components and the board.

Best regards,

 Stefan

Hi Stefan,

Therefore, we can only use EXCEL to calculate the preliminary compensation RC first, and then conduct actual measurements to confirm whether the phase margin and gain margin are OK, and adjust the most suitable compensation RC.

Hi Vin,

yes, this is the best method to ensure a stable system.

Best regards,

 Stefan

Hi Stefan,

Does TI have any documents that can help me understand how to create a Bode plot using the LM5156?

Hi Lin,

this document give as great introduction into Bode plot and loop stability:

Switch-mode power converter compensatin made easy

Best regards,

 Stefan

Hi Stefan,

I know I need to add a resistor in the feedback loop and inject a signal for measuring the Bode plot using an isolation transformer on that resistor.

But I'd like to understand how the value of the added resistor is determined.

Hi Lin,

this depends on the vector network analyzer and is typically in the range of 10 - 50 Ohm.
So please check on the used measurement tool what they recommend.

Best regards,

 Stefan